Small Power Regulated Power Supply

[teleman]

Yes,I stand corrected!It is indeed rated @500mA.(12VA /2x12v)

I can of course easily run this PSU at a reduced voltage, around +/-12-13v as suggested by another forum member(mjf).I don't really need to run any of my current projects at +/-15v at all!
besides, the max. current requirement is not more than 50-60mA at the most at the moment.then, I can of course change to a bigger "trafo" at a later juncture.

I would also need a bigger "casing", as the present one would be too small for a larger trafo!

[teleman]

lgreen, Thanks!
With regard to the 10uF caps in series with the adj.resistor,in
one of the articles @
http://www acoustica.org.uk/t/3pin_reg_notes1.html the author suggests using a "cap" twice the main frequency ie:a 100uF cap instead, for even better noise specs!

[TechGuy]

Originally posted by teleman
lgreen, Thanks!
With regard to the 10uF caps in series with the adj.resistor,one of the articles I've mentioned earlier even suggests a 100uF cap instead for better noise specs!

I'm not too sure ..so i'll read it again before confirming.

Bigger caps aren't necessarily better a reducing noise. The best caps for filtering noise are the caps with the lowest ESR. Some of the larger caps can sacrifice ESR for capacitance. To reduce noise usually a set of small Tantalum Caps is added in parallel to the electrolytic Caps.

*ESR: Equilavent Series Resistance

[teleman]

Here are two passages from the same article,which I'd like to quote directly from;

ON USING A MUCH HIGHER VALUE C2 (100uF)

1."The 317's reference voltage is based on holding an accurate 1.25v across R1.This is multiplied by (R2/R1) to get an output voltage of 1.25 (1+(R2/R1)) In effect, R2 adds POSITIVE FEEDBACK from the sum of R1 current and the internal voltage reference current! What C2 does is bypass (reduce) the positive feedback:effectively reducing 'noise gain' to the internal amplifier.

At DC, the voltage gain remains R2/R1, but to AC signals the voltage gain tends towards 0/R1 since C2 'looks like' a very small impedance at higher frequencies! THIS LOWERED NOISE GAIN is where the much better ripple rejection shown in the LM317 (and similar parts) datasheet comes from.So a BIG HINT: MAKE Z (the impedance) of C2 much smaller than R2 at twice your AC mains frequency to suppress ripple better!

2.WHY LOW-ESR CAPACITORS CAN BE A BAD THING!

An error amplifier tries to force the output up or down by comparison with an internal referance. As the error amplifier's gain falls off with increasing frequeny,so too does the regulator's accuracy and output impedance!

This rise in output impedance with frequency then just looks like an inductor-and that is how it behaves:as a synthesised inductor.

At the output of a regulator,then using a cap with too low an ESR risks forming a "peak" in the overall response, because it can resonate with this inductive output impedance!

Ladies & Gentlemen, the floor is YOUR'S!

[AndrewT]

or go the whole hog and fit an ordinary 220uF to the adjust pin of the 317.
Even better is to fit a cap//Zener to define a voltage at the adjust pin. The 317 sends an almost perfect constant current through the Zener and thus holds a very stable voltage at the adjust pin. Now that internal amplifier stands a good chance of giving a nice stable output voltage.

[Gopher]

Andrew

What jbau's results (and others I've seen - I can provide refs if necessary) seem to show is that the conventional wisdom of putting a big cap on the adj pin is wrong. There best phase response is obtained with something around 60-120nF, allthough this will significantly reduce ripple rejection, something jbau hasn't considered so far.

Another thing - since the o/p Z of an LM317 depends heavily on current draw, isn't all this fine tuning a bit pointless anyway?

Any thoughts on this?

[AndrewT]

Originally posted by Gopher
jbau's results

I have been following that thread since the beginning. I have said little, deliberately, since his philosophy is at odds with what I have come to learn. But, I listen and watch intently to give him (and others) a chance to re-educate me.
I think I am in for a long wait.

Anyone who puts in that much experimentation and time and is prepared to post the bad results along with the good results along with the unexpected/unexplainable results is a brave soul who deserves to be listened to.

[Gopher]

Andrew

I've come to respect your knowledge on matters audio, especially PSUs.

Could you provide a sketch or description showing what you think an optimised LM317 circuit looks like?

[AndrewT]

Originally posted by Gopher
Could you provide a sketch or description showing what you think an optimised LM317 circuit looks like?

I want to wait until Jbau finalises his guidance.

a brave soul who deserves to be listened to

I meant it.

[Gopher]

This seems a minefield. So many variables.

Let's see - the LM317 output Z varies with both frequency and output current (I've no info on whether the impedance curves even look the same shape for different Iout), also in/out Vdiff if jbau is correct, o/p capacitor size and ESR and Vadj capacitor size and ESR. Then there's the effect of the impedances of local Vsupply traces/decoupling caps to consider if one reg supplies several different chips. The latter of course will be different from one circuit to the next.

How on earth can one optimisation scheme fit all cases?

Oh, and we haven't even considered snubbers on the o/p of the reg yet!!

Oi vey.